The Venezuelan heath ministry released data on infant deaths that the international media reported as showing that infant mortality had increased 30% in 2016. Reuters, the BBC, NBC news among others put out headlines announcing infant mortality “soaring” or some equivalent description. I am not inclined to give the media a pass often but the uncritical reporting in this particular case was understandable given Venezuela’s economic crisis.

The data came from a report the government used to put out much more regularly – a “boletin epidemiologcio” as it is called in Spanish  The “boletin” shows a 30% increase in the number of infant deaths in 2016 compared 2015. The number of infant deaths in a year is not the same thing as an infant mortality “rate” (which is the deaths per 1000 live births) but the big problem with the news reports is seen when you look at boletin data from previous years.

For example, a 2012 boletin showed a 19.2% increase in the number of infant deaths in 2012 compared to 2011. It is very easy to show that the infant mortality rate decreased in 2012. It did not increase at all, never mind increase by 19.2% which is what you would conclude by applying the international media’s analysis of the latest boletin data.

First of all, UNICEF data for 2012 shows Venezuela’s infant mortality rate as 13.8 deaths per 1000 live births – about 2% lower than its 2011 value. Second, Venezuela had one of the fastest growing economies in the region that year (and real GDP per capita reached one of its highest levels ever thanks to several earlier years of strong growth) which allowed former president Hugo Chavez to decisively win the presidential election in 2012. His strongest base of support was always among the most vulnerable.

Mind you, UNICEF data for individual years are often projections (an epidemiologist pointed out to me) that typically take a while to get updated for short term traumatic events like Venezuela’s present economic crisis, so I am not ruling out that the infant mortality rate may have increased in 2016. However, it is clear from previous years that the boletin data for infant deaths is a far from reliable indicator of trends in the infant mortality rate.

UPDATE May 13, 2017:

An added problem is that the term “infant mortality” is often used as a shorthand for the “infant mortality rate”. I’ve done it myself – in this very blog post in fact which I have now edited for further clarity.